I'm going to preface this with a few requirements and caveats.
1) It will only work in a league that has existed for a few seasons.
2) It is very difficult to manage during the draft. Try some mock drafts with it first.
3) You need to have some Excel skills
4) It need to use Rotisserie Scoring
5) You're somewhat relying on projections and it is not a guarantee of success
6) Don't tell anyone else what you are doing
I am in a Fantasy Hoops league (the single best fantasy sport due to it's variety of numbers, but that is another argument for another thread) that has been around since the 1990's. It has been the same 6 guys every year and we have finally perfected the cateories we use. So, I looked at the average stats for the team that won each category in each season and set up a spreadsheet that would add the projected numbers of each player together. Then, the spreadsheet would compare the stat number that was required to win the category. It would then tell me what percentage of the total number needed to win the stat that I currently had. So, I could see which stat I needed more and which I had covered adequately. So, for example, after three picks last year, I found that I was already at 40% of my offensive rebounds but only at 15% of my assists needed to win the category, so I knew what to target. Finally, I set up an average of my category percentages, and the higher this got, the better. So, as I my picks were approaching, I would enter the targeted players into the spreadsheet and compare their impact on both the individual categories and my overall standing.
Now, this spreadsheet does have some problems. It told me that Danny Granger was a better fantasy player than both Kobe and Wade, but it didn't take into account that he is always hurt. Still, I'm pretty happy with the results.
This is what is called "nerding-out."